Privatising, liberalising and dividing a welfare state without affecting universality? Debunking the myths surrounding the rapid rise of private health insurance in Sweden

Abstract

The privatisation of provision and the emerging privatisation of funding, manifested in the rapid rise of private health insurance, are the most obvious signs that the universal, Swedish health system is gradually weakened. Meanwhile, the private welfare industry creates a neoliberal Newspeak where the burdening effects of the private insurance system on public healthcare are said to be unburdening, and where every step away from the principles of a universal welfare model is said to be in line with the principles of a universal welfare model. The language spoken by the private welfare industry spills over into authorities, journalists and scholars. In this article, I discuss, problematise and partially reject two research questions – Does VHI unburden the public health care system? and Are VHI holders less supportive of funding public health care? – where the authors fail to place development in the context of increased inequality and declining tax ratio, and where they use the welfare industry's definitions invented to blur the consequences of a parallel health system.

1. Introduction

Does the rapid rise of private health insurance in Sweden burden or unburden the public health care system? Does private health insurance affect the willingness to pay taxes? These are the two questions raised in an article entitled ‘Does voluntary health insurance reduce the use of and the willingness to finance public health care in Sweden?’, written by Kullberg et al. (2021).

To answer the first question, the authors investigate whether policyholders utilise the public health care system in a less degree than those without insurance. To answer the second question, the authors analyse data from a survey where policyholders are asked about their willingness to pay taxes.

In their article, the authors come to the conclusions that private health insurance unburdens the public health care system, and that policyholders are as supportive of funding public health care as non-policyholders.

To put their two questions and conclusions in a broader perspective, it must first be said that a health care system can be weakened and transformed into another system. The privatisation of provision and the emerging privatisation of funding, manifested in the rapid rise of private health insurance that is currently taken out by 15% of the adult population (Swedish Insurance, 2021), are some of the most obvious signs that the Swedish welfare model gradually loses its most fundamental characteristics: the public provision and funding of central welfare services (Pontusson, 1987; Premfors, 1991; Cox, 2004).

The rise of private health insurance is thus an example of a gradual movement towards a more liberal (Esping-Andersen, 1990) or divided welfare state (Hacker, 2002) where parts of the population choose private or semi-private welfare solutions within the hidden welfare state (Howard, 1999), while the rest remain in the visible, public system that was previously common to all citizens.

Hence, the rise of private health insurance cuts straight into decisive issues about the future funding of welfare and what type of welfare model Sweden shall be. Here, proponents of a more liberal and divided health care system have long spoken about the importance of lower taxes, increased private funding and more private health insurance (e.g. Bröms, 2004; Borg, 2009; Svanborg-Sjövall and Örtengren, 2019).

However, despite the quest for lower taxes, more private funding and more private health insurance, i.e., measures that make the welfare state less universal in every meaningful sense of the term, advocates of a more liberal and divided health system often make the strange claim that such development does not weaken the universal welfare model in any way (e.g. Morin, 2016; Pettersson Westerberg and Erlandsson, 2018).

The main reason for not defending privatisation with more straight forward arguments, for example by claiming that universality is obsolete and that health care shall be a commodity as any other, is probably that large parts of the population are still sympathetic to the universal welfare model and the Health Care Act (HSL, 2017) with its principles of health care provided on equal terms for all citizens. Cox (2004) has named the phenomenon ‘the path dependency of an idea’, i.e., that a welfare model can be so popular that its opponents feel compelled to allegedly support it even if their proposals go in a totally different direction.

In sum, it is impossible to understand the rise of private health insurance if failing to place development in a context of changing welfare models. However, this mistake is made in the article mentioned above (Kullberg, Blomqvist and Winblad, 2021), an article that partly takes off from my research (eg. Lapidus, 2017, 2019, 2020) and the public debate that I have initiated over the rapid rise of private health insurance in Sweden.

Here, I discuss and problematise Kullberg, Blomqvist and Winblad's article in which they, by asking the two questions (Does VHI unburden the public health care system? and Are VHI holders less supportive of funding public health care?), takes on the thankless task of proving that obvious steps towards a liberal welfare model are not steps towards a liberal welfare model.

2. Does VHI unburden the public health care system?

Does VHI unburden the public health care system? This is one of the two questions that Kullberg et al. (2021) try to answer in their article. Let us start by analysing this first question because the second one – Are VHI holders less supportive of funding public health care? – is a subdivision of the overarching issue of burdening or unburdening effects on public health care.

How shall burdening and unburdening be defined? Since the concepts are as multifaceted as debated, such clarification is extra important. But without further ado, Kullberg et al. (2021) use the definition put forward by representatives of the private welfare industry, namely that the more people who buy and use private health insurance, the more the public health system is unburdened (Morin, 2016; Pettersson Westerberg and Erlandsson, 2018).

Kullberg et al. (2021) first question – Does VHI unburden the public health care system? – thus boils down to an investigation of ‘whether having a VHI is associated with lower utilisation of the public health care system compared to those without private health insurance’, and their conclusion is:

Taken together, the results presented here indicate that VHI holders used public health care services to a somewhat lower extent than non-holders. This suggests that an unburdening effect of VHI might exist.

Could they have come to another conclusion? The answer is no, because, of course, policyholders use their insurance to a certain extent. Why else would they buy it? And since they buy it and use it, it automatically follows that private insurance unburdens the public health system.

What Kullberg et al. (2021) actually say is thus that the more people who leave the public health care system, the more the public health care system will prosper. All in all, that is a very static view of the consequences of a growing private insurance market within a universal health system. Let us therefore, in a more dynamic way, approach the ways in which private health insurance affects the public health system.

In fact, rather than unburdening the public health care system, private health insurance burdens the public health care system in many different ways. Here, we look into burden by definition, burden on trust, burden on access to health professionals, and burden on public finance.

3. Burden by definition

At the most general level, the rise of a parallel and half private health care system involves a very drastic burden on public health care, so drastic that it destroys the public system as we know it to paraphrase Bill Clinton's ‘end welfare as we know it’ (Soss and Schram, 2007).

Two health care systems that work according to two essentially different logics are simply not the same as one health system that works according to one and the same logic. When more and more people take out private insurance and in abandon what was the universal health system that was common to all citizens, well, then the system is no longer common and universal.

One of the most distinctive features of the Swedish welfare model was the public monopolies providing publicly funded welfare services (Pontusson, 1987; Premfors, 1991; Cox, 2004). Since the early 1990s, however, both of these distinctive features have been gradually eroded. First, we have seen substantial privatisation of the provision of health care. Second, we have seen gradual privatisation of funding of health care, manifested in the rapid rise of private health insurance.

The purpose of the Swedish, supplementary (Thomson and Mossialos, 2009) version of private health insurance is to gain quick access to health care, usually at the mainly publicly funded but privately run clinics where publicly funded patients are waiting their turn. On their websites, the private healthcare providers describe how policyholders shall proceed to jump the queue (Ersta Diakoni, 2020; Capio, 2021). Today, there are hundreds of privately run clinics where insurance patients are prioritised on non-medical grounds before the publicly funded patients (Lapidus, 2019), something that is challenging the Health Care Act (HSL, 2017) according to which healthcare must be provided ‘on equal terms for the entire population’ and where ‘those who have the greatest need for health care must be given priority’.

The Swedish health care system was built in the statutory (HSL, 2017) purpose that all citizens, regardless of class affiliation, would find content in a public system where healthcare would be provided as needed and on equal terms. It was a system that gradually and comparatively frictionlessly (Immergut, 1989) created public monopolies on the provision and funding of health care.

Today's ongoing privatisation is a movement in the opposite direction, and private provision and private funding reinforce each other in different ways. First, the emerging privatisation of funding would not have been possible without the previous privatisation of provision, which started already in the early 1990s. It is still the case that only privately run care facilities receive both policyholders and publicly funded patients, so without extensive privatisation of provision, the insurance companies would have had nowhere to send their 700,000 customers.

The insurance market is thus dependent on privatisation of provision, but private care providers also tend to become increasingly dependent on the insurance companies. It can be risky for profit-maximising actors to rely solely on the income from county councils. Further, the insurance companies pay more per patient than do the public (Lapidus, 2019), something which may lead to refusals of publicly funded patients as has happened in several US states (Decker, 2012).

Moreover, private health insurance in Sweden requires a parallel infrastructure of new actors such as insurance companies and insurance brokers, and new professions such as care coordinators and loss adjusters. It also requires technical solutions such as booking systems where private care providers and insurance companies can meet (Lapidus, 2019) to guarantee what is the main purpose of a supplementary (Thomson and Mossialos, 2009) private health insurance in Sweden, namely that policyholders get quick access to treatment (approximately two weeks for the entire care process) in comparison with the times guaranteed for the publicly funded patients (approximately six months for the entire care process) (The Swedish Agency for Health and Care Services Analysis, 2020).

The rapid rise of private health insurance creates a new type of inequality. First, the purpose of the insurance is to gain faster access to care on a non-medical basis, usually at the same clinic where other people are waiting their turn. Second, the business idea behind private health insurance is to avoid adverse selection (Cutler and Zeckhauser, 2000) by excluding certain groups and creating smaller groups of people who are not too sick, old, poor or unable to work.

However, prioritising on non-medical grounds also occurs in the workplaces. Those who do not have a benevolent employer cannot make a favourable gross wage deduction, and those who are not fully able to work are excluded from the insurance collective (Lapidus, 2019). Furthermore, the insurance companies have a profit interest in finding out the class composition in the companies, as blue-collars have a greater risk of getting hurt than white-collars.

For example, requirements for exclusion of non-healthy workers and demands for information on class composition were set when two municipal housing companies were to take out private insurance for their employees. One of them withdrew its procurement after I noticed it in the media, but the CEO of the other company clarified that employees who ‘for health reasons has specially adapted work’ shall not have the right to private health insurance (Lapidus, 2021).

The insurance market grows horizontally as more citizens take out insurance through their employers and their unions. The blue-collar unions within LO still argue that insurance policies erode public healthcare and threaten the Swedish welfare model (Johansson and Lorentzi, 2019), but the white-collar unions within Saco and TCO are increasingly offering insurance to their members, even though some of them have turned insurance companies down with the motivation that it creates A and B teams in the health care sector (Publikt, 2020).

Insurance is also growing vertically, i.e., that more care segments and increasingly highly specialised care are included in the insurances, something which in turn creates demand for new types of insurance. An example of the former is the newly built private cancer clinic Perituskliniken, and an example of the latter is the specialised cancer care insurance offered by the insurance company Alivia (Alivia, 2021).

In sum, a parallel system that requires its own infrastructure but is intertwined in the public system through full access to private but mainly publicly funded care providers, is a burden by definition on a public system that was supposed to be equal, universal and common to all citizens. This, however, is not problematised or even mentioned in the article by Kullberg et al. (2021).

4. Burden on trust

Trust in public welfare systems is fundamental for these systems to function as well as possible (Rothstein and Uslaner, 2005). But what happens when more and more people buy priority at the same clinics where others are waiting their turn? And what happens when all citizens can see that there is a semi-private system that works better than the public one, which is supposed to be run according to the motto ‘Only the best is good enough for the people’.

There are studies showing that healthcare privatisation reduces trust (eg. Cammett et al., 2015), but all in all the questions above are largely rhetorical. If ‘inequality stands at the beginning of the causal chain’ (Rothstein and Uslaner, 2005) in explaining low levels of trust, then trust will obviously be negatively affected when some get better conditions than others even in areas where the law states that conditions shall be the same for all citizens.

Meanwhile, it is the business idea of the private welfare industry to reduce citizens' trust in the public health system. First, by pointing out the public health system's deficiency in comparison with private health insurance. Second, by more or less subtle remarks that it is the shortcomings in the public health system which forces people to seek private welfare solutions.

As for the first argument, it is anything but subtle. Every insurance company puts its own product in contrast to public healthcare and states that ‘you avoid the waiting times in public care’ or ‘you do not have to wait for help and care’ (Movestic, 2021; Skandia, 2021). The insurance companies claim that ‘unfortunately public cancer care does not live up to the guidelines’ (Andersson, 2021), while launching its own concept by saying that ‘Alivia's goal is for everyone in Sweden to have the same opportunities for good care’ (Andersson, 2021) when in fact risk groups are excluded from the formal right to take out the insurance (Alivia, 2021).

The constant questioning of the public health system is spreading rapidly in society, not least to employers, trade unions and individual policyholders who need to legitimise the sometimes-criticised behaviour to buy before other citizens in the care queue. Despite the fact that Swedish health care is highly ranked internationally (Barber et al., 2017), these loops of negative policy feedback (Mettler, 2002) can gradually change the connotation of public health care from something that citizens feel proud of to the exact opposite, much as the concept of welfare has changed throughout American history (Soss and Schram, 2007).

In the end, even critics of private health insurance begin to blame its emergence on the shortcomings of public health care, and here we find the argument that ‘private health insurance is a symptom of the failure of public health care’ (Waltersson Grönvall, 2020). This argument is more subtle than just pointing to the badness of public health in general. It conveys the image that if only public care had worked a little better, a private insurance industry would never have emerged.

But that is simply not true. A public system can never compete with a private but state-subsidised system that is aimed only at relatively healthy, young, rich, able-bodied and thus less costly citizens. Not even during the heyday of the Swedish welfare model, the public sector would have stood a chance against the insurance system, if such was allowed to emerge in the same way as today.

In sum, trust in the public health care system is affected in many ways by the rapid rise of private health insurance. However, this is not mentioned by Kullberg et al. (2021).

5. Burden on access to health professionals

Lack of staff in the public health system can become an even greater problem in the near future. Andersson (2019) concludes that gradual tax increases are necessary if the quality is to be maintained, and similar reasoning can be found in reports from the Swedish Association of Local Authorities and Regions (SKL 2010) and the National Institute of Economic Research (Konjunkturinstitutet, 2018). Here, it is often pointed out that money is not enough if there is no staff to employ. ‘The problem is the great shortage of staff’, says Annika Wallenskog, chief economist at the Swedish Association of Local Authorities and Regions (Ahmadi, 2021).

Where, then, are all the health professionals needed in the public health system? There are many answers to that question (e.g. Taylor, 2020), for example that too few choose to become health professionals and that too many leave the profession for various reasons.

A third answer is that health professionals are drawn into the private insurance industry. A country has only a certain set of staff. Every doctor who treats someone who has jumped the queue could, at that very moment, have treated someone who is waiting their turn in the public system.

As we have seen, this leakage of care staff usually takes place at clinics that also receive publicly funded patients. The private caregivers thus let the staff work with both categories of patients (Anderzzon, 2017).

Further, health professionals are employed by the insurance industry not only as practising doctors and nurses. The parallel infrastructure requires staff in other positions, for example as care coordinators and loss adjusters, and there are lots of such adverts for nurses who are in great shortage in the public health system (e.g. Studentjob, 2021):

Finally, health professionals are needed when insurance companies want to make sure that private providers do not perform unnecessary surgical procedures just to make extra money. The insurance companies Länsförsäkringar, Euro Accident and SEB Pension & Försäkring have teams of medical experts to review cases where it is suspected that providers have operated on patients without any such operations being justified (Länsförsäkringar, 2016). These are regular battles between insurance companies and private providers and their respective teams of medical experts (eg. Försth et al., 2020).

6. Burden on public finance

Private health insurance is so dependent on the public sector that we should talk about semi-private or pseudo-private rather than private health insurance. On the one hand, it is an indirect state sponsorship where the insurance industry is a rent-seeking (Tollison, 1982) free-rider (Grossman and Hart, 1980) on public healthcare. First, health professionals who are not only educated and trained by the public sector, but who also get large parts of their many years of specialist practice paid for by the public (Swedish Medical Association, 2014), can receive customers with private health insurance. Second, policyholders often come to premises where rent and other practicalities are already paid for by the public sector via the agreements that they have concluded with the private care provider. Third, the entire insurance market is based on getting access to private but mainly publicly funded clinics (Lapidus, 2019).

In addition to the above and under-explored form of indirect state sponsorship, we also have the direct state support through tax breaks that is so typical of the hidden welfare state (Howard, 1999).

In the Swedish case, the direct state sponsorship is based on gross salary deductions where the state covers half the cost of the insurance. That was possible due to the lack of benefit taxation. In 2018, however, a law on benefit taxation was enacted (Government Bill, 2018). The law abolished the subsidies directed mainly to the healthcare consumption of the higher social classes.

The centre-right wing political parties were against the law, and the Confederation of Swedish Enterprise had advocated even greater tax breaks than those that applied before it was enacted (Confederation of Swedish Enterprise, 2016). Further, editorial writers from a number of newspapers started to talk about ‘triple taxation’ (e.g. Gudmundson, 2017) despite the fact that tax on benefits is customary in Sweden.

However, there were those who went even further. The business organisation Swedish Insurance and the insurance company Skandia wrote letters to the Swedish Tax Agency and arranged meetings with its legal experts on two occasions. The results were above expectations. The Tax Agency reinterpreted and changed its original position (Swedish Tax Agency, 2018) to a new one (Swedish Tax Agency, 2019) where 40% of private health insurance are exempt from taxation.

Swedish Insurance spread the good news to its customers: ‘At the request of Swedish Insurance, the Swedish Tax Agency has clarified that the taxable benefit of a health insurance is set at 60%’ (Swedish Insurance, 2019).

7. Burden in many other ways

The rapid rise of private health insurance is a burden on the public health care system in a number of other ways which, if it was not for lack of space, would also be worth its own headings.

First, health care does not function like other markets (Arrow, 1978), which is why semi-private healthcare systems tend to be more expensive in many different ways (e.g. Reinhardt et al., 2004; Bodenheimer, 2005). In short, private care providers want to sell as much health care as possible to policyholders who want as much health care as possible in order to get value for their insurance money. Further, there are high transaction costs (Williamson, 1981) due to control mechanisms established because of the lack of trust in and between private, profit-maximising providers. And the more the market is deregulated, the more costly regulations and re-regulation these deregulations seem to require (e.g. Majone, 1997).

Second, a culture of anxiety is created around health care. Policyholders in the new healthcare market compare with each other and do not want to fall behind (Frank, 2013) in the healthcare race; a competitive situation that insurance companies take advantage of for example by offering bronze, silver, gold and platinum insurance (Euro Accident, 2021; WellCare, 2021). This kind of status contest has a negative impact not only on those at the bottom of the inequality pyramid, but also on those at the top of it (Wilkinson and Pickett, 2010).

Third, it is not true that more resources remain for those who stay with public health care, one of the most frequent status quo arguments used to defend the parallel health care system (e.g. Morin, 2016; Erlandsson, 2019). The healthcare budget is updated continuously, and it reflects the resources that are considered necessary. Politicians, constantly pressured by austerity requirements (Blyth, 2013), are not late to cut costs if possible. The more people who leave public health care, the less resources politicians have to target to the public health care budget.

Fourth, it has opinion-related consequences that it is primarily the most vocal social groups that sign private health insurance. They get quick access and no longer have to worry about problems in the public system. As a result, a group of citizens who might have put a lot pressure on politicians to improve public health care, disappear. This can be grateful for politicians who seek to avoid blame (Weaver, 1986) for cut-downs or, which is more common, for the policy drift (Béland et al., 2016) of neglecting to update healthcare budgets to the current economic situation.

8. Are VHI holders less supportive of funding public health care?

One of the two conclusions in the article by Kullberg et al. (2021) is that the willingness to pay taxes is not affected among those who sign up for private health insurance. But how to find out if the willingness to pay taxes increases or decreases? As we shall see, there are difficulties measuring tax willingness by asking people, and it is even more difficult to distinguish those with private health insurance from all other groups.

Given the difficulties, it is extra important to relate to general development and study the facts on the ground, for example how taxes develop over time. The Swedish tax ratio dropped dramatically over the past 20 years. In 2000, it was 48.6%. In 2020, it was 42.6% (Statistics Sweden, 2021).

Important reasons for the declining tax ratio are the abolishment of a number of taxes for the wealthy (e.g. inheritance tax 2005, wealth tax 2007, property tax 2008, austerity tax 2020), while various types of tax breaks were introduced in several sectors, such as the gross salary deduction for private health insurance and the so-called Rut deduction that can be used for a long and ever-increasing range of services (Regeringen, 2020) aimed primarily at the upper-income strata of the population (Lapidus, 2019).

In terms of money, there are tax breaks that cause even larger holes in the state budget, such as the interest deduction and above all the earned income tax credit (Järliden Bergström et al., 2020) which was introduced by the centre-right wing government 2006–2014.

It happened a lot during the years of the centre-right wing government. But in the middle of that period of dramatic tax cuts, 2011, there is a frequently cited – so also by Kullberg et al. (2021) – study that shows that the willingness to pay taxes is as strong as ever (Svallfors, 2011). And not just the popular support for high taxes for that matter, but in fact the centre-right wing party-political support for universal welfare as a whole:

What seems to have taken place in the last few years is that since their main party – the Moderates [right-wing] – has embraced the core aspects of the welfare state, even the higher salaried and the self-employed have increasingly become supporters of a collective welfare state. The Social Democratic Party may be in dire straits electorally, but the social democratic welfare state is more popular than ever (Svallfors, 2011).

It is quite remarkable to draw these conclusions in the middle of a period that sought radical changes in the welfare model (e.g. Etzler, 2013). Just a few years later, however, Svallfors changed his mind and saw many radical changes in the Swedish welfare sector (Svallfors, 2016).

But let us stick to the previous study (Svallfors, 2011) and take a look at the measures of the willingness to pay taxes, where the same questions have been asked on different occasions (1986, 1992, 1997, 2002, 2010), for example ‘Do you think that the amount of tax money used for the following purposes should be increased, remain the same or decreased?’.

Svallfors (2011) finds ‘a large degree of stability’ over time when it comes to overall attitudes, even though individual sectors jump up and down in (in terms of healthcare, for example, there is a clear decline between 2002 and 2010).

But what do the answers actually mean if several taxes have been abolished during the same period, and if the tax ratio has dropped dramatically?

For example, those who answered ‘Remain the same’ in 2010 can be assumed to advocate lower taxes than those who answered ‘Remain the same’ the last time the question was asked. If Svallfors finds that as many people answer ‘Remain the same’ in 2010 as in 2002, and if taxes fell during that period, then the willingness to pay taxes have probably fallen rather than remained the same.

Svallfors does not take falling tax rates and falling tax ratio into account and the same goes for Kullberg et al. (2021) who, using data from one single point in time (SOM-Institute, 2016) (this was the first time the SOM-institute included questions on private health insurance), try to separate policyholders from the rest of the population. As long as they separate policyholders from the rest of the population in general, the authors find a clearly lower willingness to pay taxes among policyholders. But they also want to control for sex, age, self-assessed health, household income, education and political orientation.

Separating policyholders from people who are exactly the same in all other aspects runs into many problems. First, there is a heated public debate where policyholders' solidarity with the public health system is questioned and where representatives of the private welfare industry, in order to legitimise the parallel system, insists that the support for public care is not at all affected.

In this situation, can it really be assumed that policyholders honestly answer questions about willingness to pay taxes? Or do they rather have an interest to appear as good citizens who want to contribute as much as possible, just as the private welfare industry tells them?

Here it might be the same problem as with many other opinion polls, such as ‘the Bradley effect’ where white voters in California did not want to say that they would vote for a white candidate for fear of being perceived as racist (Payne, 2010). Or ‘the shy Tory factor’, where British Conservatives did not want to say that they were to vote for the Conservative Party (Fisher and Lewis-Beck, 2016).

Second, no one assumes that private health insurance alone and in itself has a huge impact on willingness to pay taxes. Rather, it is but one of many factors and phenomena and trends in a neoliberal development that gradually leads to reduced willingness to pay taxes and falling tax ratios. What can be expected is thus, if it had been possible to measure correctly, marginal differences between policyholders and non-policyholders with the exact same political orientation.

Third, the rise of private health insurance reduces trust in public care in a way that do not only affect policyholders and their willingness to pay taxes, but all people. A generally reduced willingness to pay tax makes it even more difficult to distinguish the policyholders from everybody else, and especially from those with the same political orientation.

Kullberg et al. (2021) refer to several studies that find decreased willingness to pay taxes among policyholders compared with the rest of the population (eg. Hall and Preston, 1998; Costa-Font and Jofre-Bonet, 2008). The only study that is said to show the opposite (Martinussen and Magnussen, 2019) is in fact hesitant and concludes that:

Our results mean we cannot rule out that a market for collective parallel coverage PHI through employers may undermine the support for the publicly financed health care.

Despite all difficulties and doubts, Kullberg et al. (2021) conclude that the rapid rise of private health insurance has not ‘altered the support for the principle that health care should be financed in a solidaristic manner by all members of society’.

Given the difficulties of measuring tax willingness by asking policyholders about it, given the development of a constantly declining tax ratio in a situation where the tax ratio must increase if public healthcare is to maintain its quality and status, given that liberal welfare models (the direction in which Sweden is gradually moving) are low-tax countries (OECD, 2020), and given that private health insurance is the opposite of direct tax-financed healthcare, it would have been desirable for Kullberg et al. (2021) to place their study in a larger, societal context and also to present a theoretical explanation for the conclusion that willingness to pay taxes is unaffected by private health insurance on a massive scale.

The problem is that there is no such explanation. On the contrary, publicly provided and funded welfare for all people is increasingly questioned as more actors are drawn into the privatisation logic (e.g. Hacker, 2002), for example the Swedish white-collar unions who gradually leave behind the doubts that existed when leading representatives said things as ‘There is a moral problem when the legislation says that care shall be provided due to needs while we, parallelly, establish a system where you actually can buy healthcare’ (Mörtvik, 2011), and where the same representatives warned about a ‘vicious circle’ (Propper, 2000) which now seems to be closed: ‘The more unions that provide [private health insurance], the greater the competitive disadvantage for the others, which makes development self-generating’ (Mörtvik, 2011).

In this context, mention must be made of ‘The paradox of redistribution’ (Korpi and Palme, 1998) according to which it is easier to get support from the upper classes for universal welfare solutions than for measures aimed only at the poor, and that ‘programs for the poor are poor programs’ (Van de Walle and Nead, 1995). Thus, programmes aimed only at the poor ‘stimulate program exit among the middle classes and increase the demand for private insurance’ (Korpi and Palme, 1998).

Recent research (e.g. Busemeyer and Iversen, 2020) indicates that the ‘broad cross-class support for the universalistic welfare state that is implied by this scholarship [e.g. Korpi and Palme, 1998] can break down in the transition to a world with viable and high-quality private alternatives’. What Busemeyer and Iversen mean is that support for universal solutions among the upper social classes is falling apart ‘as we move from a before private alternatives (BPA) world to an after private alternatives (APA) world’, and they write:

But once private alternatives are introduced, support for continued public provision is undermined by the fact that high-income citizens opt out of public schemes and in turn become more supportive of a selective rather than a universalist model of the welfare state.

Where we possibly differ is that I think Busemeyer and Iversen (2020) and Korpi and Palme (1998) basically talk about the same thing, only that the former focus on the top of the pyramid (introduction of private alternatives) while the latter focuses on the bottom of it (means-testing).

I propose a synthesis. Concepts such as ‘parallel societies’ and ‘means-testing’ usually refer to poor and vulnerable groups, but in fact they should just as often refer to the top of society. More often than not, they are two sides of the same coin. State-subsidised private health insurance is a kind of means-testing for the rich, but also a kind of means-testing even of those who have to make do with an increasingly poorly functioning public system. Although the latter group do not have to prove old or sick or incapable of working enough to be entitled to public healthcare, they are demonstrably too old or sick or incapable of working to have access to the parallel system. In sum, welfare programmes for the rich are also poor programmes.

9. Conclusion

The rapid rise of private health insurance burdens the public health system by definition. Two healthcare systems that work according to two essentially different logics are not the same as one healthcare system that works according to one and the same logic. When more and more people take out private health insurance and abandon the universal system common to all citizens, then it is no longer a universal system common to all citizens.

There are a number of burdening factors that are more specific than the general burden by definition. For example, public healthcare loses health professionals to the insurance industry not only when they are busy treating those who are prioritised on non-medical grounds, but also because the parallel system requires an infrastructure where health professionals work as care coordinators, claims adjusters, supervisors and more.

Further, private health insurance reduces trust in public healthcare. It is the business idea of the insurance companies to portray public healthcare in a bad light, and it is legitimising for policyholders to act in the same manner. The loops of policy feedback gradually change the picture of, and trust in, one of the world's best public health systems.

In addition, private health insurance is not that private since it implicates direct costs for the public sector by generous tax breaks, and above all indirect costs when policyholders in violation of the Health Care Act are welcomed to private but publicly funded providers where they freeride on publicly trained and paid care staff and publicly paid premises and equipment.

Among all neoliberal and tax-reducing trends, it is difficult or even impossible to isolate the negative impact of private health insurance on the willingness to pay taxes. Here, one must make qualitative reasoning based on comparisons with other countries and based on facts on the ground, not least how the tax ratio develops during the era of privatisation. In Sweden, the tax ratio is drastically falling when it should rather rise to maintain the quality of public welfare services.

It is a thankless task to argue that obvious steps towards a liberal welfare model are not obvious steps towards a liberal welfare model. Those who try must put the issue in context and explain how it differs from other neoliberal mechanisms that spur the trend towards increasing income gaps and tax cuts that primarily benefit the already rich. They must also avoid using the static definitions invented by the private welfare industry.